Introduction to Mainframe (z/os) Network Management

Transcription

1 Introduction to Mainframe (z/os) Network Management Monday, August 10, 1:45-2:45 Session Dean Butler

2 Agenda What is network management? Why manage the network on z/os? z/os networking Network monitoring tools and technologies Best practices 2

3 What is network management? FCAPS ISO standard model Accounting / Admin Goal: Distribute resources optimally and fairly among network subscribers. Measure utilization of all important network resources. Analyze results. Regulate, bill, or charge users. Fault management Goal: Keep network operational, minimize downtime. Find and correct network problems. Identify future potential problems and prevent from occurring or reoccurring. Performance management Goal: Measure and make network performance data available to optimize performance. Gather, analyze, and threshold performance data. Configuration management Goal: Monitor network and system configuration to track and manage changes. Monitor and control network operation. Coordinate hardware and programming changes. Maintain an inventory. Security management Goal: control access to assets in the network Manage network authentication, authorization, and auditing. Maintain data security with authentication and encryption. 3

4 Network management architecture Management system Network management protocol Management entity Network Agent Management database Agent Management database Agent Management database Managed devices 4

5 Agenda What is network management? Why manage the network on z/os? z/os networking Network monitoring tools and technologies Best practices 5

6 Why manage the network on z/os? Why the network? A well-managed network will allow you to get on with running your company. Prevent problems Work efficiently Maintain security Stay up to date Why on z/os? In spite of perception that z/os networking doesn t break, SLA s are at risk: Application changes Hybrid networks BYOD Device failures Network congestion Network systems programmer: A critical application is broken. We all get on a bridge call. Everyone says My stuff is okay. It must be the network. I need to be able to say it is not the network or at least not my part of the network. 6

7 Agenda What is network management? Why manage the network on z/os? z/os networking Network monitoring tools and technologies Best practices 7

8 System z network connectivity Via external network Via OSA adapter z/os LPAR Linux LPAR z/os LPAR Linux LPAR 1 OSA K or Mbits OSA Express with Queued Direct I/O (QDIO) The HiperSockets technology is based on the QDIO technology from the OSA Express adapter. For HiperSockets, it is referred to as internal QDIO - or iqdio for short. z/os LPAR 3 Capacity and speed increase. Via HiperSockets HiperSockets GBytes OSA Linux LPAR OSA Gbits A HiperSockets network looks like an internal LAN 2 HiperSockets allows operating system images on the same processor complex to exchange IP traffic virtually at memory speed. 8

9 z/os Communications Server Integrated Services Provide common services within CS Network attachment Storage management High Performance Data Transfer TCP/IP and SNA integration TN3270 Network access Internal optimizations Enterprise Extender Standard TCP/IP applications Multi-protocol Solutions Sockets (TCP/IP) applications Unix services offers z/os users access to a wide range of UNIX-based applications over IP networks SNA applications SNA applications are supported over SNA or IP networks z/os Communications Server SNA APIs SNA, APPN, HPR Applications Communications Storage Manager Enterprise Extender Network Attachment DLCs Sockets APIs TCP/IP 9

10 Enterprise Extender Allows use of IP network for SNA sessions o To an SNA application, the IP network looks like an HPR link o To the IP network, EE is just a UDP application SNA traffic is sent as UDP datagrams (on ports ) over the IP network Supported by many platforms: o z/os o CS/Linux, CS/AIX, CS/Windows PComm o i5/os o Cisco SNASw o Microsoft HIS Typically isolates SNA footprints to the "outside" of the network. IP Backbone TN3270, or Web client IBM SNA Clients System z Enterprise Servers IBM HPR IBM Cisco SNASw, Communications Server for Linux or other IBM ICN SNA Network TCP sessions/routes SNA routes for SNA sessions EE routes for SNA sessions EE allows enablement of IP applications and convergence on a single network transport while preserving SNA application and endpoint investment. 10

11 Designed by the magnificent Overbymeister Design ed by the "Chief" Lynn Long How z/os communicates with the outside world Enable IBM and ISV network management products to provide differentiated value on System z Enable network connectivity with differentiated z/os Qualities of Service Performance Security High Availability z/os Network Mgt Products: Tivoli (Netview, OMEGAMON, zsecure, TSOM), ISVs Network Management Interfaces CICS, IMS, DB2, MQ, WAS, WebSphere middleware, IM software, Batch, Rational, Tivoli, ISV, RYO Transactions Data Serving Web Services z/os Communications Server Network Communications APIs TCP/IP and SNA z/os Differentiation and Value Add Security Network Layer Protocol Stacks TCP/IP V4 TCP/IP V6 SNA Network Attachment Drivers Applications High Availability Performance Network Applications and Services FTP, TN3270 SMTP, Pagent, SNMP, z/os Comm Server Configuration Assistant Web Based GUI Part of z/os Management Facility Enables all key z/os workload to communicate with the outside world Every request for service and response flows through Comm Server Provides standard suite of TCP/IP applications but with close integration into z/os Enables network connectivity using standards based network protocols IPv4: Current internet protocol, support for latest standards IPv6: Emerging next generation Internet protocol SNA: Maintain existing customer application investment Network Adapters IPv4 SNA IPv6 11

12 Typical data center network z/os Network 12

13 Agenda What is network management? Why manage the network on z/os? z/os networking Network monitoring tools and technologies Best practices 13

14 Network diagnostic commands PING NETSTAT TRACERTE DISPLAY NSLOOKUP VARY DROP 14

15 Network diagnostic commands PING Test connectivity to a host or device on an IP network ICMP echo request / echo response Measures round trip time TRACERTE Shows most likely path to an IP device Series of ICMP or UDP packets Uses TTL to identify network hops in path Measures round trip time to each hop 15

16 Network diagnostic commands NSLOOKUP Query the Domain Name System (DNS) Translate host name to/from IP address DROP Terminate a TCP/IP socket endpoint TCP listener, TCP connection, UDP endpoint Requires CONTROL access to MVS.VARY.TCPIP.DROP 16

17 Network diagnostic commands NETSTAT Display information about the local host, including TCP/IP configuration, connections, network clients, gateways, and devices Drop connections (access to MVS.VARY.TCPIP.DROP) 17

18 Network diagnostic commands z/os operator commands: DISPLAY TCPIP Display information about the local TCP/IP or TELNET address spaces VARY TCPIP Control some functions of local TCP/IP and TELNET address spaces 18

19 What is SNMP? Simple Network Management Protocol One of the most widely accepted protocols to manage and monitor network elements. UDP application Each SNMP agent maintains an information database describing the managed device parameters SNMP manager uses this database to request information from the SNMP agents SNMP agent sends some information as traps Network management protocol SNMP manager Basic commands: Get, Get next, Get bulk Set Traps Inform Response Managed devices SNMP Agent MIB SNMP Agent MIB Network SNMP Agent MIB 19

20 What is SNMP? An Object Identifier (OID) identifies a managed object. For example, the OID in RFC1213 for "sysdescr" is SNMP packet flows: SNMP Manager SNMP Agent 20

21 System Management Facilities (SMF) Standardized method for writing out records of activity to a file (or data set to use a z/os term) Full "instrumentation" of all baseline activities running on an LPAR, including I/O, network activity, software usage, error conditions, processor utilization, etc. SMF forms the basis for many monitoring and automation utilities Each SMF record has a numbered type (IBM: 1-127, others: 128+) Installations have great control over how much or how little SMF data to collect TCP/IP statistics are captured in SMF 109, 118,

22 Designed by the magnificent Overbymeister Designe d by the "Chief" Lynn Long z/os Communication Server Network Management Interfaces (NMI) API to obtain information about TCP/IP and VTAM resources: Alternative for most SNMP, NETSTAT, and DISPLAY output Efficient! Methods of collection: Callable APIs Event information Real-time packet trace z/os Network Mgt Products: Tivoli (Netview, OMEGAMON,zSecure,TSOM), ISVs Network Management Interfaces CICS, IMS, DB2, MQ, WAS, WebSphere middleware, IM software, Batch, Rational, Tivoli, ISV, RYO Transactions z/os Communications Server Network Communications APIs TCP/IP and SNA z/os Differentiation and Value Add Security Network Layer Protocol Stacks TCP/IP V4 TCP/IP V6 SNA Network Adapters Data Serving Web Services High Availability Performance Network Attachment Drivers Applications Network Applications and Services FTP, TN3270 SMTP, Pagent, SNMP, z/os Comm Server Configuration Assistant Web Based GUI Part of z/os Management Facility Most modern network management products on z/os use the NMI to collect data. IPv4 IPv6 SNA 22

23 Agenda What is network management? Why manage the network on z/os? z/os networking Network monitoring tools and technologies Best practices 23

24 Network management process Plan / Design: - Build history - Baseline - Trend analysis - Capacity planning - Procurement - Topology design Plan / design Implement / deploy Implement / Deploy: - Install and configure - Address management - Adds, moves, changes - Security - Accounting/billing - Assets/inventory - User management - Data management Operate / maintain Operate / Maintain: - Define thresholds - Monitor exceptions - Notify - Correlate - Isolate problems - Troubleshoot - Bypass/resolve - Validate and report 24

25 Baseline Your Environment Become familiar with your network resources. Know what is: Normal Abnormal but uninteresting Abnormal and interesting Be part of your change control system Review periodically!!! 25

26 Problem isolation z/os Comm Server: TCP/IP IP layer TCP layer UDP layer TCP/IP address space VTAM Buffer pools EE HPR VTAM address space z/os Comm Server Hardware or network Application Application: Applications TCP Listeners TCP Connections UDP Endpoints Hardware or network: OSA ports OSA channel HyperSockets Interfaces Data center network Intranet Internet 26

27 Isolate problem Routing Table CICS, IMS, DB2, WAS, FTP, TN3270,... SNA Sockets Sockets Sockets sessions HPR APPN sessions TCP Layer UDP Layer Connections Listeners UDP Endpoints Enterprise Extender IP Layer Interfaces OSA Express Hipersockets ICMP Application TCP/IP or VTAM Hardware or network 27

28 Protocol indicators Applications Sockets TCP/UDP Sockets IP ICMP Sockets IP Stack Gateway/Devices Interfaces OSA-Express OSA-Express TCP/IP Input Discards Output Discards UDP Discard UDP Input Errors UDP No Port OSA Fragments, jabber, length error, CRC, alignment Unknown IP Frames Interfaces Inbound Packets discarded, in error Outbound Packets discarded, in error Utilization Transmission Rates Unknown IP Frames 28

29 Resource constraint indicators TCP/IP Applications Sockets TCP/UDP Sockets IP ICMP Sockets IP Stack Gateway/Devices Interfaces OSA-Express OSA-Express CPU Percentage CSA Allocated, In Use Authorized Private Storage Allocated, In Use ECSA storage max, allocated, in use, pools Datagrams Discarded Backlog Connections Rejected UDP Datagrams Discarded HPR Throughput rate allowed Unacknowledged Buffers high water mark Interfaces Receive/Transmit Bandwidth utilization OSA PCI Utilization Processor Utilization Missed Packets VTAM CPU Percentage CSA Allocated, Allowed, In Use 29

30 Indirect indicators TCP Applications Sockets TCP/UDP Sockets Sockets Segments Retransmitted Response Time Fragmentation and/or reassembly Out of Order Segments Remote Window Size Frequency TCP Keep-Alive Drops IP ICMP IP Stack Gateway/Devices Interfaces HPR Out of Sequence Buffers Packet Retransmission Rate Path Switches Response Time Variance Smoothed Round Trip Time OSA-Express OSA-Express TN3270 Average IP Response Time and variance Average SNA Response Time and variance 30

31 Scenario: User reports batch FTP failures The setting: A company relies on batch FTP to copy files between a mainframe at headquarters and each of its retail stores every night (local store time). Sales and inventory data is uploaded and product and pricing changes are downloaded to the stores. One morning, a systems administrator notices that some of the files have not been updated for days. He reports the problem to the IT help desk. The problem is routed to the mainframe networks systems programmer. 31

32 User reports batch FTP failures USS Storage FTP TCP/IP FTP Sessions FTP Transfers FTP Transfers Start with checking current activity: FTP transfers & FTP sessions 32

33 Check TCP/IP Stack Retransmits network congestion Out of Order routing issue or network congestion Fragmentation MTU size Discards resource constraints Timeouts connectivity UDP input errors attack UDP discards no application running Sockets Sockets TCP/ UDP Routing Table CSM & Buffers High storage utilization - could indicate network congestion This can result in requests backing up in storage IP Interfaces OSA-Express OSA-Express 33

34 Check OSA and Interfaces OSA Online Status Configuration Microcode Level Utilization Transmission Rates Unknown IP Frames By LPARS By Ports Interfaces Packet Errors Bandwidth Utilization MTU Size 34

35 Check Applications and Connections Applications: Accepting connections? Rate, Backlog, Rejections Last activity time Response time Retransmissions Transmit / Receive Rates Out of order segments Connections: Start time/duration Response Time Retransmissions Transmit / Receive Rates Out of order segments 35

36 Network is OK, then where is the problem? Check job logs, SYSLOG, and syslogd for clues. Dataset access issue? FTP session login failure? 36

37 Introduction to Mainframe (z/os) Network Management Session Dean Butler 37